on avoiding temporal biases in ranking systems

looking at the imdb top250 list .. it is remarkable to see the bias towards recent releases .. for e.g the departed has already made it to the top 200 .. somehow the mechanism doesnt seem to be very robust since there is an inherently positive bias towards recent happenings .. wonder if there are standard statistical techniques to avoid such biases -- but again a classic counter-example is the shawshank redemption which was a huge failure but is still very high on the list due to subsequent dvd successes.